Nursing Home Bed Supply and Medicare Inpatient Utilization: 1981-86.

摘要

The study investigates the substitutability of nursing home care for alternative dimensions of hospital care. The hypotheses of the study are that increases in the ratio of Medicare or Medicaid certified nursing home days to the elderly population in an area will not only reduce average Medicare length of stay but will also tend to reduce the rate of hospital admissions and readmissions. Data sources included Medicare claims data aggregated to the county level for 1981 through 1986, the Medicare/Medicaid Automated Certification System (MMACS) (for certified nursing home bed counts) and the Area Resource File. Linear regression and minimum logit chi-square models were estimated in both cross-section form and in a fixed-effect formulation that holds persistent county-level characteristics constant over time. While increased availability of nursing home beds was found to reduce the average length of hospital stays, it also appeared to stimulate rather than reduce hospital admissions for aged beneficiaries living in urban areas and to increase the likelihood of rehospitalization, contrary to the authors' hypothesis. The relation between nursing home beds and hospital use is positive even after controlling for persistent county-specific effects. While the relationship could result from a violation of certain assumptions discussed, the finding is sufficiently robust to suggest that it is a real phenomenon.